Formalities
The course consists of six meetings held on Mondays or Tuesdays (depending on your study group, check the schedule here) at 18:10 on Kantemirovskaya. Each session includes a lecture followed by a practical coding workshop. Students are expected to attend classes, participate actively in discussions, and engage in coding activities. The content of the final meeting (on December 8th or 16th, the gap is due to the holiday on November 4th) is flexible and can be adjusted based on student requests.
Your mark for the course consists of:
- homework assignments: 60% (5 assignments in total; the average of 3 highest scores)
- final exam: 40% (individual or group project)
All assignments and projects should be submitted via email to ensure timely review and feedback. The sections below outline the details of both grading components.
Home assignments
There will be 5 homework assignments throughout the course, but only your best 3 scores will be used to calculate your final grade. This means you can choose to complete all the homework or just 3 without it affecting your final mark. For the homework component, I will simply average your 3 highest grades. However, I strongly recommend reviewing each assignment, as completing them will better prepare you for the final project.
Homework assignments are due before the start of the next class (the exact time is 18:10, either Monday or Tuesday, each week). If you submit an assignment within the following week, you can still earn up to 8 points.
Exam
The exam can be completed either individually or in groups of up to 4 students. Your task is to conduct a small research project and report your findings in a paper of approximately 4–5 pages (including visualizations, tables, and bibliography). These length guidelines are flexible, so feel free to write more or less if needed. The exam will consist of the following elements:
1. Collect network data
You are free to choose your context — whether it’s networks of friends, Hollywood actors, or another topic of interest. If you are working individually, you may use pre-existing datasets from GitHub or other data archives. However, ensure you don’t simply copy someone else’s work entirely. Replicating a study (without access to the original code) can be a good approach, though. I would not decrease the mark for analyzing already collected data, though you can get additional 2 points if you do collect the data yourself.
2. Describe your data
Provide a brief overview of your dataset and highlight its key aspects. Include a readable visualization of the network you are analyzing and compute and discuss its descriptive properties (e.g., density, diameter). Be sure to consider the context of your data: how ties are formed, how actors are selected, and any other relevant details.
3. Pose research questions
Formulate 2–3 research questions (preferably interconnected) and explain why these questions are important. If you have any hypothesEs and supporting literature, include them in your report. Note that a larger team need either more questions or questions which are difficult to work with alone.
4. Describe your workflow
Discuss the materials you consulted, any ideas that came to you during the course, and how long the analysis took. Include methodological considerations and reflections on the choices you made throughout the project.
5. Present and discuss the results
Present your findings — for example, this could be in the form of a table with network values. However, it is important to also provide interpretations of these results, explaining what they mean in the context of your research.
Don’t hesitate to explore and experiment with the tools, as the goal is to learn through hands-on experience. Feel free to change the structure (e.g., data collection and description, research questions, methodology/inspiration/etc., results) if you feel it does not align with the type of work you are doing. The project format is flexible to suit different research interests and approaches.
In an ideal scenario, you should start thinking about your final project around the second meeting. Having ideas early will allow you to benefit more from class discussions and ask questions that specifically relate to your project.
The main purpose of the exam is to assess whether you are familiar with the basic logic of network analysis and whether you can apply some of the methods covered in the course. While the quality of your analysis is important, the focus will also be on your effort, reflection, and interpretation. Even if your technical analysis is not perfect, what matters most is your understanding of the principles of network analysis and your ability to critically reflect on your process. Network analysis can be challenging, and working with real data often presents unexpected difficulties, so the evaluation will not be overly strict. The deadline for the exam report is 19th December (by the end of the day) for both groups.
Important note: last year some students decided to present their work during the last seminar, and I freed from writing this report (they got their exam marks for their in-class presentation). If there would be any volunteers, we can arrange similar activities during the final sessions this year.
Evaluation
The following grading criteria apply to both homework assignments and the final exam:
- “Excellent” (9–10)
The student demonstrates deep knowledge and advanced skills, exceeding the materials discussed in the class and/or incorporating additional relevant resources. Coding procedures are thoroughly commented on, all tasks are completed without errors, and the work is properly structured.
- “Excellent” (8)
The student shows a strong understanding of the topic. While minor mistakes may be present, they do not significantly affect the results or interpretations. The work is well-structured and includes detailed explanations of the analytical procedures.
- “Good” (6–7)
The student responds correctly to most tasks but may provide some misleading interpretations and/or occasional errors that affect the results and interpretations. The work is clearly structured with appropriate comments.
- “Satisfactory” (4–5)
The student addresses about half of the assigned tasks and/or makes significant errors. Results and interpretations lack depth, and the work is not detailed enough to fully explain the analysis. Formatting is unclear.
- “Fail” (0–3)
The student fails to demonstrate knowledge of the relevant topics. Most tasks are either incorrect or not completed, and interpretations are brief or missing, with little to no reference to appropriate concepts or methods.
Extensions and late submissions
If you are unable to submit a homework assignment or take the exam on time due to a valid reason (e.g., illness, conference participation, or other extenuating circumstances), please notify me as soon as possible. You may receive up to 2 additional weeks to submit a homework assignment after the general deadline, and up to 1 additional week to complete the exam.
Documentation (e.g., a medical certificate) is required to confirm the reason for the delay. Please ensure to communicate any issues before the deadline whenever possible.
Software requirements
The primary tool for this course will be R and RStudio. R offers excellent packages for social network analysis, and it arguably provides better coverage of these methods compared to Python. I will guide you through learning how to effectively use these packages.
While R is powerful for analysis, it’s not the most intuitive tool for producing network visualizations. Due to that, we will explore Gephi during our third meeting, as it is a highly effective tool for vizualization. Additionally, I encourage you to explore online tools like Cosmograph and GraphCommons for inspiration; these platforms also allow you to create visually appealing graphs.